Conversion Rate Using A/B Testing

7 replies
Hello Everyone, I am a newbie to learning A/B Testing, I want to know that Why is the use of A/B testing and conversion optimization important for a business?
#a or b #conversion #rate
  • Profile picture of the author TobiMDD
    hmm isn't that kinda obvious? I mean more conversions can lead to more sales and more sales mean more money and this is what a business wants..
    {{ DiscussionBoard.errors[11699731].message }}
  • Profile picture of the author dave_hermansen
    If you go to pizza place #2 and find that it is better than pizza place #1, you likely won't go back to pizza place #1. Next, you try pizza place #3 and see if it is better than pizza place #2. If so, you now ditch pizza place #2 and pizza place #3 is your go-to place for delicious pie. And so on and so on ...

    Same thing with split testing a website. You're trying to find the combination that brings the most visitors and conversions. Not all pizza is the same. Not all content, placement and colors are the same. Some are better than others and you won't know what is best until you make a lot of changes one at a time.
    Signature - The #1 place to buy & sell websites!
    We help sellers get the MAXIMUM amount for their websites and all buyers know that these sites are 100% vetted.
    {{ DiscussionBoard.errors[11699817].message }}
  • Profile picture of the author Freemon Barber
    It is so you won't be in the dark just guessing what works.
    For a Landing page:

    1.)You write the copy on that page to entice visitors to opt-in (usually an email address).

    2.)Then you make another page exactly the same... BUT you make one SMALL change..(ex... a different headline)

    3.) When both pages are ready, you send an equal amount of traffic to both pages (at least 100 visitors but ideally 1,000) and see which page gave you the most email sign-ups.

    The difference may be huge. You don't know until you A/B test.

    4.) You stop using the page that lost, and create another page just like the winner but again, with one small change... and do it all over again.

    5.)Repeat that process until it's as good as you can possibly get it-----on a lead page shoot for 40-50%. conversion rate opt-in rate.

    6.)Once you have a real winner converting very high you can properly estimate how much it will cost you to get a lead and be pretty confident about it. You will know alot more about your money---how much you'll be spending and ballpark how much you'll be earning for each person you send to the page.

    7.) You will continue to use the winning page that is working for as long as you can until it no longer works. Almost nothing works forever. But by then, you'll be able to confidently do A/B tests on any other campaigns.

    Test that first and then you can move on to testing sales page headlines and body copy and email subject lines, etc, etc...

    P.S. If you're using solo ads on udimi or something or emails in general to get people to your opt-in page---- depending on your autoresponder service--- you may easily be able to test a couple different email subject lines and see which one helps get you the most clicks to your opt-in page that you are testing, ---BUT, so you get a strong handle on it, I would start with the landing page. It takes patience, but it could mean a lot more money for you, so remember: Only change one thing at a time so you know what worked and you won't get confused and have to do it all over again. --Best of Luck!
    {{ DiscussionBoard.errors[11701086].message }}
  • Profile picture of the author Jason Kanigan
    Originally Posted by Kennyanderson View Post

    Hello Everyone, I am a newbie to learning A/B Testing, I want to know that Why is the use of A/B testing and conversion optimization important for a business?
    The purpose is to test and find performance increases.

    So you put a funnel up, run traffic to a conversion tool and see how well this combination performs. In my training I go a lot deeper than this, but that's the essence of it.

    After a few months of testing you have a baseline. We'll call this A.

    Now you begin to wonder, based on things you've heard from your traffic and ideas you've encountered in the marketplace, whether a different conversion tool--call to action, format, hook etc.--would perform better than A. So you create a new funnel, which we'll call B.

    The smart thing to do while testing is run 75% of traffic to A, 25% to B. This way you keep making money while you test.

    Most people horrifically underestimate the volume of traffic required to do real A/B split testing. Most is invalid. 100 people in your sample is not enough. You need thousands and it takes awhile. Otherwise, your results are not statistically relevant (or truthful).

    So you run both funnels for awhile, until you have a statistically relevant result. Then you can say whether funnel A or funnel B converts better. At that point you stop sending traffic to the worse performing one and send it all to the other... and then that becomes the new A. Eventually you'll get an idea and want to create a new B, and the testing cycle continues.

    At least, that's the theory. In a dozen years of online marketing what I've seen most marketers do is one A/B test, not run it long enough to get truly statistically relevant results, drop the one that looks like it's performing worse, and stop. Then they run traffic to A until the end of time--or whenever the mojo runs out and the funnel stops converting (probably because they ran out of qualified traffic, and developed ad fatigue for the audience by showing them the same thing over and over and over again). Then maybe they get their butt in gear and create new content.

    But in the meantime, they probably ran off to work on something else that their shiny object syndrome waved at them
    {{ DiscussionBoard.errors[11701178].message }}
  • Profile picture of the author Randall Magwood
    It's important so that you can cut out the losers and continue to roll out with the winners.

    If you're advertising on Google Ads, Microsoft Ads, and doing some banner advertising on site #3... you will want to find out which one(s) should get the bulk of your money.

    If you invest $200 into each of them, and find that Google Ads outperforms the other 2 platforms by alot... you can effectively cut out the other 2 platforms, and put that other $400 into Google Ads - thus tripling your good results.

    But you'll never know this if you advertise blindly and dont test. This is the general gist of the idea.
    {{ DiscussionBoard.errors[11701490].message }}
  • Profile picture of the author Klara Pelhe
    To be honest, A/B testing is really a great thing and unique opportunity to accurately test and compare results you are achieving. Without this, it can be really hard to compare metrics between different campaigns and see what is working and what is not. I highly suggest that you use this often and set various variables on various campaigns to test many creatives, placements, budgets, and so on. This can be used on social media, Google ads, or on email marketing, so it's really great to have these testings included.
    {{ DiscussionBoard.errors[11701719].message }}
  • Profile picture of the author Rivin Mithun
    Well, just like you want to treat your loved ones with best burger. But to choose which one is best you have to try them both and have to decide which tastes good. So as like this A/B testing is for business to choose what will be the best decision for you to convert
    {{ DiscussionBoard.errors[11719801].message }}

Trending Topics